Related Experiment Video
Updated: Oct 25, 2025
10:27
In Silico Identification and Characterization of circRNAs During Host-Pathogen Interactions
Published on: October 21, 2022
1.7K
A systematic evaluation of the computational tools for lncRNA identification
Hansi Zheng1, Amlan Talukder1, Xiaoman Li2
1Department of Computer Science, University of Central Florida, Orlando, FL, USA.
Briefings in Bioinformatics
|August 9, 2021
Summary
Deep learning tools excel at identifying long non-coding RNAs (lncRNAs), but their performance drops when considering cell-type-specific transcripts. Peptide features offer minimal accuracy improvements for lncRNA identification.
Area of Science:
- Computational biology
- Genomics
- Bioinformatics
Background:
- Long non-coding RNAs (lncRNAs) play crucial roles in cellular functions.
- Accurate computational identification of lncRNAs is essential for their study.
- Existing lncRNA identification tools lack systematic performance evaluation and feature importance analysis.
Purpose of the Study:
- To systematically evaluate the performance of 17 computational lncRNA identification tools.
- To assess the importance of various features used by these tools.
- To provide guidance for selecting optimal tools and features for lncRNA identification.
Main Methods:
- Performance assessment of 17 lncRNA identification tools.
- Evaluation across multiple common benchmark datasets.
- Investigation into the contribution of different features, including peptide features.
Main Results:
- Deep learning-based tools demonstrated superior performance in general lncRNA identification.
- Peptide features were found to have a limited impact on overall tool accuracy.
- A significant performance decline was observed for all tools when evaluated on cell-type-specific transcript data, with deep learning tools losing their advantage.
Conclusions:
- Deep learning approaches are highly effective for general lncRNA identification.
- Feature selection, particularly regarding peptide features, requires careful consideration.
- The context-specificity of RNA transcripts significantly impacts the performance of lncRNA identification tools, necessitating specialized approaches for cell-type-specific analyses.
